Allogeneic Haematopoietic Stem-Cell Transplantation With

ORIGINAL ARTICLE Allogeneic haematopoietic stem-cell transplantation with reduced intensity conditioning for advanced stage Hodgkin’s lymphoma in Sweden: high incidence of post transplant lymphoproliferative disorder

J-E Johansson1, M Remberger2, VLj Lazarevic3, H Hallbo¨o¨k4, A Wahlin5, E Kimby6, G Juliusson3, H Omar6 and H Ha¨gglund6

1Department of Hematology and Coagulation, Sahlgrenska University Hospital, Go¨teborg, Sweden; 2Division of Clinical Immunology and Transfusion Medicine, Karolinska Institutet, Stockholm, Sweden; 3Department of Hematology, Ska˚ne University Hospital, Lund University, Lund, Sweden; 4Department of Hematology, Uppsala University Hospital, Uppsala, Sweden; 5Department of Hematology, Cancer Centre, University Hospital, Umea˚, Sweden and 6Hematology Center, Karolinska University Hospital Huddinge, Stockholm, Sweden

Allogeneic transplantation after reduced intensity con- compared with conventional dose salvage chemotherapy— ditioning (allo-RIC) is a treatment option for patients with can successfully treat patients with chemosensitive relapsed Hodgkin’s lymphoma (HL) relapsing after autologous Hodgkin’s lymphoma (HL),1,2 no prospective, randomised transplantation. In all, 23 adult patients with HL under- studies have examined how to manage patients with relapse went allo-RIC in Sweden between 2000 and 2007. The or progression after autologous transplant. Several retro- median number of previous treatment lines was five and 20 spective reports have evaluated allogeneic transplantation patients (87%) were previously autografted. TRM at 100 after full intensity conditioning3–6 and several more recent days and at 1 year was 13 and 22% respectively. Acute studies have evaluated allogeneic transplantation after GVHD grades II–IV developed in 7 out of 23 patients reduced intensity conditioning (allo-RIC).7–15 In these (30%) and chronic GVHD in 10 out of 20 patients at risk trials, the selected patients were in an advanced disease (50%). The OS and EFS at three years was 59 and 27%, stage and heavily pre-treated and both TRM and the respectively. Four patients (17%) developed post trans- relapse rates were high. The present trial, the first national plant lymphoproliferative disease (PTLD) after a median compilation study, summarises experiences of allogeneic time of 55 days (range 38–95); two of these patients later transplantation after reduced intensity conditioning in died. The study confirmed that allo-RIC is feasible, adult HL patients in Sweden. but associated with a substantial relapse rate: only 20% of the patients were still alive 7 years after the transplant. A finding of high incidence of PTLD needs to be Patients and methods confirmed in a larger trial that includes patients with non-HL and CLL. The patients and their donors were identified using the Bone Marrow Transplantation (2011) 46, 870–875; Swedish transplant centres’ UPN-lists. Data were collected doi:10.1038/bmt.2010.238; published online 18 October 2010 locally at each participating centre from patient files, local Keywords: allogeneic; transplantation; Hodgkin’s lym- registries and the European Group for Blood and Marrow phoma; post transplant lymphoproliferative disorder; Transplantation (EBMT) database for all adult patients reduced intensity conditioning (X18 years) who had been treated with allogeneic transplantation after reduced intensity conditioning due to HL. In total, 23 patients from five centres were transplanted between January 2000 and December 2007. Data were available for all 23 transplanted patients. One Introduction patient was reported in a previously published EBMT study,16 but data on the remaining 22 patients had not been Although two randomised studies have demonstrated that used in earlier studies. The study protocol was approved by high-dose therapy with autologous stem cell rescue— the Regional Ethical Review Board in Goteborg and patients gave informed consent before transplant. The median number of treatment lines before the Correspondence: Dr J-E Johansson, Department of Hematology and allogeneic transplant was five (range 2–8), and 20 patients Coagulation, Sahlgrenska University Hospital, S-413 45 Go¨teborg, (87%) had previously undergone an autologous transplant. Sweden. E-mail: [email protected] The median time from diagnosis to the allograft was 53 Received 14 June 2010; revised and accepted 23 August 2010; published months (range 17–162 months). Three patients (13%) were online 18 October 2010 chemorefractory at the time of the allograft. The donors Allogeneic transplantation for Hodgkin’s lymphoma J-E Johansson et al 871 were HLA-identical related in three cases (13%), HLA-A, Conditioning regimen -B and –DR identical unrelated in 14 cases (61%) and The conditioning was considered reduced or non-myeloa- mismatched unrelated in six cases (26%). Patient blative according to the consensus criteria proposed by the characteristics are described in Table 1. Regimen-Related Toxicity Working Committee of the Center for International Blood and Marrow Transplant 12,17,18 Table 1 Patient characteristics of the 23 patients included Research, criteria that has been used in other studies. The conditioning regimens used are described in Table 1. Characteristic Measurement

Age (Years) 36 (19–49) GVHD prophylaxis Men 13 (57) Women 10 (43) In two cases, CYA alone was used and in the remaining 21 cases CYA together with a short course of MTX was used. Donor Anti-thymocyte globuline was used for in vivo T-cell Matched related 3 (13) depletion before transplant if the donor was unrelated. Matched unrelated 14 (61) Mismatched unrelated 6 (26) Stem cell source CMV serology (recipient/donor) À/À 5 (22) In 17 cases (74%), PBSCs were used and in the remaining À/+ 5 (22) six cases (26%) BM was used. The median CD34 cell dose +/À 8 (35) count was 5.4 Â 106/kg (range 2.1–12.4). +/+ 5 (22)

EBV serology (recipient/donor) Statistics +/+ 12 (52) The analysis was performed in February 2010. The À/+ 2 (9) +/À 1 (4) probabilities of overall survival and relapse-free survival unknown/+ 4 (17) were estimated using the method developed by Kaplan– +/unknown 3 (13) Meier and compared with the log-rank test.19 The incidence unknown/unknown 1 (4) of GVHD, TRM and relapse were estimated non-para- metrically. Patients were censored at the time of death, Female donor—male recipient 4 (17) Time from diagnosis to allograft (months) 53 (17–162) relapse, or last follow-up. Relapse and non-relapse No. of previous treatment lines 5 (2–8) mortality are competing events. Their incidence rates were Previously autografted 20 (87) estimated using a non-parametric estimator of cumulative Time from autograft to allograft (months) 22 (9–63) incidence curves.20 Predictive analyses for GVHD, TRM Histological subtype and relapse were based on the proportional hazard model Nodular sclerosis 22 (96) for subdistribution of competing risk. Univariate and Lymphocyte rich 1 (4) multivariate analyses were then performed using Gray’s test and the proportional sub-distribution hazard regres- Status at allo transplantation sion model developed by Fine and Gray.21 All tests were CR or CRu 2 (9) PR 18 (78) two-sided. The type I error rate was ﬁxed at 0.05 for factors Refractory 3 (13) potentially associated with time-to-event outcomes. Analyses were performed using the cmprsk package Conditioning (developed by Gray, June 2001), Splus 6.2 software (S-plus Flu/Mel 14 (61) Flu/Bu 4 (17) 6.2, Insightful, Seattle, WA, USA) and Statistica software Flu/Cy/TBI 3 (13) (Statistica, StatSoft, Tulsa, OK, USA). Flu/Cy 2 (9)

Antilymphocyte antibodies used Yes 20 (87) Results No 3 (13) Engraftment and GVHD Stem cell source All 23 patients were engrafted. The median time to reach an BM 6 (26) 9 PBSC 17 (74) ANC of more than 0.5 Â 10 /L and to achieve a platelet count of more than 20 Â 109/L was 17 and 16 days, CD34 dose ( Â 106/kg) 5.4 (2.1–12.4) respectively (Table 2). One patient’s platelet count was never below 20 Â 109/L during the transplant course. Acute GVHD prophylaxis CyA/Mtx 21 (91) GVHD developed in 10 out of 23 patients (43%). Chronic CyA 2 (9) GVHD (cGVHD) developed in 10 out of 20 patients at risk (50%). In all, 7 out of the 10 patients (70%) who developed Abbreviations: Flu/Bu ¼ fludarabine 150 mg/m2, busulphan 8 mg/kg; cGVHD had no previous acute GVHD (de novo cGVHD). Flu/Cy ¼ fludarabine 150 mg/m2, CY 60 mg/kg; Flu/Cy/TBI ¼ fludarabine The extent of cGVHD was limited in two cases and 180 mg/m2, CY 60 mg/kg, TBI 3 Gy Â 2; Flu/Mel ¼ fludarabine 150 mg/m2, melphalan 140 mg/m2. extensive in eight cases. The 1-year and 3-year cumulative Data are no. of individuals (%) or median (range). Percentage may exceed incidence of cGVHD was 45% (Figure 1). Transplant 100 owing to rounding. outcome is summarised in Table 2.

Bone Marrow Transplantation Allogeneic transplantation for Hodgkin’s lymphoma J-E Johansson et al 872 Table 2 Transplantation outcomes (n ¼ 23) 1.0 100 days: 13% 1 year: 22% Follow-up all patients (months) 25 (1–92) 0.8 Follow-up of survivors (months) 46 (3–92)

Engraftment 0.6 Neutrophils (days to 40.5 Â 109/L) 17 (11–28) Platelets (days to 420 Â 109/L) 16 (11–35) TRM 0.4

Acute GVHD 0.2 Grades 0–I 16 (70) Grades II–IV 7 (30) 0.0 0 365 730 1095 Day of onset of acute GVHD 22 (16–59) Patients at risk of chronic GVHD 20 Days after HSCT

Chronic GVHD Figure 2 Transplant-related mortality. None 13 (65) Limited 2 (10) Extensive 8 (40) Four patients (17%) developed post transplant lymphoproliferative disease (PTLD) after a median time of 55 days Day of onset of chronic GVHD 174 (100–362) (range 38–95), two of these patients died. The patients who De novo chronic GVHD 7 (70) developed PTLD (one male and three females) were all Patients at risk of CMV-infection/disease 18 transplanted from unrelated donors and had acute GVHD. CMV-infection 9 (50) Details of the patients are summarised in Table 3. Day of onset of CMV-infection 32 (20–42) In all, 9 out of 18 patients who were at risk (donor and/or recipient who were CMV-sero-positive) developed CMV CMV disease 2 (11) infection after a median time of 32 days (range 20–42 days). Day of onset of disease 39 (35–42) Two of these patients developed non-fatal CMV disease. PTLD 4 (17) Day of onset of PTLD 55 (38–95) Disease relapse and progression Non-relapse mortality (percent) 100 days 13 In all, 12 patients relapsed or progressed after a median 1 year 22 time of 8.7 months (range 1–39 months) after transplant; 3 years 22 four of these patients were still alive after a median follow- up time of 45 months (range 24–51 months). The Causes of death cumulative incidence of relapse/progression was 57% at 3 GVHD 2 (15) Multi organ failure 1 (8) years (Figure 3). PTLD 2 (15) Relapse 8 (62) Donor lymphocyte infusion (DLI) Abbreviation: PTLD ¼ post transplant lymphoproliferative disease. Nine patients were treated with DLI, two before clinical Data are no. of individuals (%) or median (range). Percentage may exceed 100 owing to rounding. De novo chronic GVHD: chronic GVHD without relapse and seven as a part of the salvage regimen. DLI prior acute GVHD. only was given to four out of seven patients who relapsed. DLI combined with chemotherapy was given to the remaining three patients who relapsed. Four out of the 1.0 seven patients (57%) who were given DLI as a part of the salvage regimen responded and were still alive after a 0.8 median follow-up time from DLI of 19 months (range 2– 42). Five patients who relapsed were not treated with DLI 0.6 45% mainly because of early and proliferative relapses or 0.4 previous extensive cGVHD. None of these patients are alive. Details of the 12 patients who relapsed are

Chronic GVHD 0.2 summarised in Table 4.

0.0 0 365 730 1095 1460 1825 Survival Days after HSCT Ten patients were still alive after a median follow-up time of 48 months (range 3–243 months). The cumulative Figure 1 The cumulative incidence of cGVHD. proportions of patients surviving at 1, 3 and 5 years after transplant were 62, 54 and 41%, respectively. The 5-year progression-free survival was 22% (Figure 4). GVHD Transplant-related mortality and morbidity (acute or chronic), chemosensitive disease or any other Five patients died of transplant-related causes. TRM at 100 factor did not signiﬁcantly affect OS or PFS (data days and at 3 years was 13 and 22%, respectively (Figure 2). not shown).

Bone Marrow Transplantation Allogeneic transplantation for Hodgkin’s lymphoma J-E Johansson et al 873 Table 3 Characteristics of the four patients who developed post transplant-lymphoproliferative disease

Patient PAD Donor ATG EBV- EBV- aGVHD cGVHD Onset PAD PTLD Local Treatment Outcome serology serology PTLD patient donor

M/40 NS Mismatched Yes Negative Positive II Nil d+95 Lymphocyte Colon Ganciclovir Dead of URD inﬁltration, liver IVIG PTLD d +101 EBER+, EBNA+ rituximab Â 1 F/37 NS Mismatched Yes Positive Positive I NA d+38 B cell lymphoma, Lung liver Rituximab Â 1 Dead of URD EBER+, EBNA+, lymph- ganciclovir PTLD d +51 LMP1+ nodes F/30 NS Matched Yes Positive Positive I Nil d+52 Polymorph PTLD, Lymph- Rituximab Â 4 Resolution of URD Lambda monoclonal nodes EBV-CTL PTLD Alive B cell population, IVIG in CR d +1466 EBER+ F/29 NS Matched Yes Positive Not IV Ext d+57 EBV-PCR-positive Lung liver Rituximab Â 4 Resolution of URD tested on d +48, PAD spleen PTLD Dead of not performed, lymph- GVHD d +227 FNA nodes inconclusive

Abbreviations: ATG ¼ Anti-thymocyte globuline; CTL ¼ cytotoxic T-lymphocytes; EBER ¼ Epstein-Barr encoded RNA (in situ hybridisation); EBNA ¼ Epstein-Barr nuclear Ag (immunohistochemistry); FNA ¼ ﬁne needle aspiration; LMP ¼ EBV latent membrane protein 1 (immunohistochemistry); NS ¼ nodular sclerosis; PAD ¼ preliminary anatomic diagnosis; PTLD ¼ post transplant lymphoproliferative disease; URD ¼ unrelated donor.

1.0 1y 38% graft-vs-tumour effects have investigated allo-RIC in 2y 53% relapsed HL. These studies are looking at the disease 3y 57% 0.8 response to DLI for residual masses, mixed chimerism, or 63% documented disease progression post transplant. This 0.6 antitumor effect appears to be related to (and at the cost of) cGVHD in most but not all cases.9 In our study, 7 out 0.4 of 12 relapsed patients were treated with DLI. Three 0.2 patients also had additional chemotherapy, making the true

Relapse/progression effect of this form of immunotherapy difficult to determine 0.0 with confidence. The five patients who were not given DLI, 0 365 730 1095 1460 1825 progressed and died shortly after the relapse. The majority Days after HSCT of these patients were considered to have a far too proliferative relapse and/or previously too much of Figure 3 The cumulative incidence of relapse/progression. cGVHD to be eligible for DLI. Thus, the effect of DLI seems to be modest in proliferative relapse, but DLI may have a role to play in patients with early, not very Discussion proliferative, relapse or as prophylaxis against manifest relapse in patients with mixed chimerism with previous, The present study is the first national comprehensive report none, or mild GVHD. In the present trial, cGVHD did not summarising the results of allo-SCT for HL in Sweden. As have any influence on the relapse rate possibly owing to in previously published reports, the patients in our cohort small numbers, as in larger studies a graft-vs-tumour effect were in an advanced disease stage and heavily pre-treated has been demonstrated.16 before allotransplant—87% of the patients were previously A novel and surprising finding in the present trial was the autografted. The study confirmed that reduced intensity high incidence of PTLD. Four patients (17%) developed conditioning in this situation was associated with a PTLD, two of whom died. The incidence of PTLD after substantial relapse rate: only 20% of the patients were still allogeneic transplantation has been shown to be between alive 7 years after the transplant. A previous study from the 0.2 and 8.1% depending on the number of risk factors.23,24 EBMT presented the results of allotransplantation in It is true that a substantial part of the patients in the advanced HL patients who were transplanted between present trial was burdened with one or two risk factors for 1999 and 2001.16 In the present study, only one of the PTLD, such as the use of T-cell antibodies (83% of patients was transplanted during that period, so the two patients) and the fact that 26% of the donors were reports do not overlap. mismatched; nevertheless, it is unlikely that these facts On the basis of the results from non-randomised trials, it explain more than a doubling of the expected incidence of may be concluded that allo-RIC in advanced HL is feasible PTLD. In this context, it is important to note that in the and possibly superior to alternative salvage therapy in this reports identifying risk factors for developing PTLD after subset of patients.14,22 However, the long-term results have allogeneic transplantation, patients with a lymphoma been quite discouraging, mainly because of high relapse diagnosis have been excluded owing to the difficulty of rates. In our study, 52% of the patients relapsed, a finding separating PTLD from relapse of lymphoma. Accordingly, that agrees with earlier reports. Recent studies exploring the true incidence of PTLD after allogeneic transplantation

Bone Marrow Transplantation Allogeneic transplantation for Hodgkin’s lymphoma J-E Johansson et al 874 Table 4 Characteristics of patients and treatment of patients who relapsed (n ¼ 12) Sex/ Donor DLI DLI Chemo aGVHD/ DLI before/after Allo to relapse Relapse to DLI Status at follow-up Age doses cGVHD cGVHD (months) (months) (months after allo)

F/28 URD No 0 Yes 0/Nil — 12 — 53/Dead M/39 URD Yes 1 Yes II/Ext After 18 60 81/Dead M/39 URD Yes 1 No 0/Ext Before 9 0 46/Dead M/49 RD No 0 Yes II/Ext — 6 — 20/Dead F/46 RD No 0 Yes 0/Ext — 14 — 67/Dead M/24 URD Yes 3 No 0/Lim Before 4 0.5 9/Dead F/23 URD No 0 Yes 0/Nil — 3 — 6/Dead M/39 URD No 0 Yes 0/Nil — 9 — 20/Dead M/27 URD Yes 2 No 0/Ext After 7 1 51/Alive F/19 URD Yes 1 No 0/Nil — 9 2 50/Alive M/36 URD Yes 5 Yes 0/Ext After 16 10 days 40/Alive F/32 URD Yes 3 Yes 0/Nil — 8 2 24/Alive

Abbreviations: DLI ¼ donor lymphocyte infusion; RD ¼ related donor; URD ¼ unrelated donor.

a OS further controlled studies are warranted to identify the 1.0 patients who would gain by a transplant and to determine when in the treatment course allo-RIC should be intro- 1year 69% duced. Thankfully, the subsets of patients with HL that 0.8 3year 59% 5year 42% may be considered for such treatment are small and therefore, it is necessary that large organisations accom- 0.6 plish these trials. The finding of the unexpected high incidence of PTLD might be a chance finding with small numbers and needs to be confirmed in a larger trial that 0.4 also includes other lymphomas than HL.

0.2 Conﬂict of interest Cumulative proportion surviving

0.0 0 365 730 1095 1460 1825 2190 2555 2920 3285 The authors declare no conﬂict of interest. Days after HSCT

b PFS 1.0 References

1year 46% 1 Linch DC, Winfield D, Goldstone AH, Moir D, Hancock B, 0.8 3year 27% McMillan A et al. Dose intensification with autologous bone- 5year 22% marrow transplantation in relapsed and resistant Hodgkin’s disease: results of a BNLI randomised trial. Lancet 1993; 341: 0.6 1051–1054. 2 Schmitz N, Pfistner B, Sextro M, Sieber M, Carella AM, Haenel M et al. Aggressive conventional chemotherapy 0.4 compared with high-dose chemotherapy with autologous haemopoietic stem-cell transplantation for relapsed chemosensitive Hodgkin’s disease: a randomised trial. Lancet 2002;

Progression-free survival 0.2 359: 2065–2071. 3 Akpek G, Ambinder RF, Piantadosi S, Abrams RA, Brodsky RA, Vogelsang GB et al. Long-term results of blood and 0.0 marrow transplantation for Hodgkin’s lymphoma. J Clin 0 365 730 1095 1460 1825 2190 2555 2920 3285 Oncol 2001; 19: 4314–4321. Days after HSCT 4 Anderson JE, Litzow MR, Appelbaum FR, Schoch G, Fisher LD, Buckner CD et al. Allogeneic, syngeneic, and autologous Figure 4 OS and PFS. marrow transplantation for Hodgkin’s disease: the 21-year Seattle experience. J Clin Oncol 1993; 11: 2342–2350. for lymphoma is unknown and therefore, it would be of 5 Gajewski JL, Phillips GL, Sobocinski KA, Armitage JO, Gale RP, Champlin RE et al. Bone marrow transplants from interest to more closely investigate the development of HLA-identical siblings in advanced Hodgkin’s disease. J Clin PTLD in patients undergoing an allogeneic transplant for a Oncol 1996; 14: 572–578. lymphoma diagnosis. 6 Milpied N, Fielding AK, Pearce RM, Ernst P, Goldstone AH. In summary, the present trial conﬁrms that allo-RIC is a Allogeneic bone marrow transplant is not better than treatment option for patients with advanced HL, but autologous transplant for patients with relapsed Hodgkin’s

Bone Marrow Transplantation Allogeneic transplantation for Hodgkin’s lymphoma J-E Johansson et al 875 disease. European Group for Blood and Bone Marrow 15 Sarina B, Castagna L, Farina L, Patriarca F, Benedetti F, Transplantation. J Clin Oncol 1996; 14: 1291–1296. Carella AM et al. Allogeneic transplantation improves the 7 Burroughs LM, O’Donnell PV, Sandmaier BM, Storer BE, overall and progression-free survival of Hodgkin lymphoma Luznik L, Symons HJ et al. Comparison of outcomes of HLA- patients relapsing after autologous transplantation: a retro- matched related, unrelated, or HLA-haploidentical related spective study based on the time of HLA typing and donor hematopoietic cell transplantation following nonmyeloablative availability. Blood 2010; 115: 3671–3677. conditioning for relapsed or refractory Hodgkin lymphoma. 16 Sureda A, Robinson S, Canals C, Carella AM, Boogaerts MA, Biol Blood Marrow Transplant 2008; 14: 1279–1287. Caballero D et al. Reduced-intensity conditioning compared 8 Anderlini P, Saliba R, Acholonu S, Okoroji GJ, Donato M, with conventional allogeneic stem-cell transplantation in Giralt S et al. Reduced-intensity allogeneic stem cell trans- relapsed or refractory Hodgkin’s lymphoma: an analysis from plantation in relapsed and refractory Hodgkin’s disease: low the Lymphoma Working Party of the European Group for transplant-related mortality and impact of intensity of condi- Blood and Marrow Transplantation. J Clin Oncol 2008; 26: tioning regimen. Bone Marrow Transplant 2005; 35: 943–951. 455–462. 9 Peggs KS, Hunter A, Chopra R, Parker A, Mahendra P, 17 Storb RF, Champlin R, Riddell SR, Murata M, Bryant S, Milligan D et al. Clinical evidence of a graft-versus-Hodgkin’s- Warren EH. Non-myeloablative transplants for malignant lymphoma effect after reduced-intensity allogeneic transplan- disease. Hematology Am Soc Hematol Educ Program 2001, tation. Lancet 2005; 365: 1934–1941. 375–391. 10 Alvarez I, Sureda A, Caballero MD, Urbano-Ispizua A, 18 Martino R, Iacobelli S, Brand R, Jansen T, van BA, Finke J Ribera JM, Canales M et al. Nonmyeloablative stem cell et al. Retrospective comparison of reduced-intensity condition- transplantation is an effective therapy for refractory or ing and conventional high-dose conditioning for allogeneic relapsed hodgkin lymphoma: results of a spanish prospective hematopoietic stem cell transplantation using HLA-identical cooperative protocol. Biol Blood Marrow Transplant 2006; 12: sibling donors in myelodysplastic syndromes. Blood 2006; 108: 172–183. 836–846. 11 Anderlini P, Saliba R, Acholonu S, Giralt SA, Andersson B, 19 Kaplan E, Meier P. Nonparametric estimation from incom- Ueno NT et al. Fludarabine-melphalan as a preparative plete observations. J Am Stat Assoc 1958; 53: 457–481. regimen for reduced-intensity conditioning allogeneic stem cell 20 Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation transplantation in relapsed and refractory Hodgkin’s lympho- of failure probabilities in the presence of competing risks: ma: the updated M.D. Anderson Cancer Center experience. new representations of old estimators. Stat Med 1999; 18: Haematologica 2008; 93: 257–264. 695–706. 12 Devetten MP, Hari PN, Carreras J, Logan BR, van BK, 21 Fine J, Gray R. Proportional hazard model for the sub- Bredeson CN et al. Unrelated donor reduced-intensity distribution of competing risks. J Amer Stat Assoc 1999; 94: allogeneic hematopoietic stem cell transplantation for relapsed 496–509. and refractory Hodgkin lymphoma. Biol Blood Marrow 22 Thomson KJ, Peggs KS, Smith P, Cavet J, Hunter A, Parker A Transplant 2009; 15: 109–117. et al. Superiority of reduced-intensity allogeneic transplanta- 13 Sureda A, Robinson S, Canals C, Carella AM, Boogaerts MA, tion over conventional treatment for relapse of Hodgkin’s Caballero D et al. Reduced-intensity conditioning compared lymphoma following autologous stem cell transplantation. with conventional allogeneic stem-cell transplantation in Bone Marrow Transplant 2008; 41: 765–770. relapsed or refractory Hodgkin’s lymphoma: an analysis from 23 Curtis RE, Travis LB, Rowlings PA, Socie G, Kingma DW, the Lymphoma Working Party of the European Group for Banks PM et al. Risk of lymphoproliferative disorders after Blood and Marrow Transplantation. J Clin Oncol 2008; 26: bone marrow transplantation: a multi-institutional study. 455–462. Blood 1999; 94: 2208–2216. 14 Castagna L, Sarina B, Todisco E, Magagnoli M, Balzarotti M, 24 Landgren O, Gilbert ES, Rizzo JD, Socie G, Banks PM, Bramanti S et al. Allogeneic stem cell transplantation Sobocinski KA et al. Risk factors for lymphoproliferative compared with chemotherapy for poor-risk Hodgkin lympho- disorders after allogeneic hematopoietic cell transplantation. ma. Biol Blood Marrow Transplant 2009; 15: 432–438. Blood 2009; 113: 4992–5001.

Bone Marrow Transplantation